expect_equivalent() gives misleading error message

[dlindelof]

When comparing two objects that are numerically different and have different names, I expect expect_equivalent() to complain about the numeric differences and ignore the names. Instead it reports a difference in the names:

> expect_equivalent(c(foo = 0), 1)
Error: c(foo = 0) not equivalent to 1.
names for target but not for current
> expect_equivalent(c(foo = 0), 0)

[krlmlr]

Works as expected for character values already:

> expect_equivalent(c(foo = "0"), "1")
Error: c(foo = "0") not equivalent to "1".
1/1 mismatches
x[1]: "0"
y[1]: "1"